Since D&D dropped the declaration phase in favor of cyclic initiative, it has let one round blur into the next without any phases that trigger between rounds.
Still, each round represents a short span of time that gives every combatant a chance to act. The end of a round and start of another makes a natural break to add phases or to switch gameplay. D&D doesn’t do any of these things, but other tabletop roleplaying games do, and RPGs have a long tradition of borrowing the best ideas from each other.
Retreat phase
Side initiative makes running away from a fight easier. In games like original D&D with side initiative, DMs could add deadly foes to their dungeons without fearing that players would get locked in a fight they couldn’t win. The party could run and likely escape. Defeated monsters could also run and possibly escape. In modern D&D, when monsters flee, the session typically stalls on a prolonged chase where the party still kills every foe.
Even without side initiative, a modern game could capture the same benefits by adding a retreat rule that switches the game’s usual turn order to side initiative. Retreat would work like this: At the end of a round, either side can declare a retreat and act together before the other side gets their actions. This enables the same sort of coordinated retreat that side initiative allowed. A full implementation of this idea might need a rule to prevent players from declaring a retreat and instead coordinating an attack.
Dialog phase
In movies, comic books, and tales by the fire, fight scenes often pause to the villain can explain their plan, offer to rule the empire side-by-side, or to just gloat. Meanwhile, heroes explain the mistakes that will lead to the villain’s fall. Roleplaying games need a dialog phase between rounds that gives characters a way to pause the action for roleplaying talk without anyone wondering why someone would waste six seconds talking while everyone else takes swings.
Making conditions land simultaneously
In D&D and games like it, effects and conditions end at different times, often at the end of another creature’s turn. Meanwhile, I struggle to keep track of what ends when. Judging from my years of playing, I’m not alone.
Could a roleplaying game reduce the burden of tracking when conditions expire by simply grouping the ends and save-to-ends at the end of each round? Such a rule would help everyone track when a particular effect expires.
However, such a process would create a problem: If a creature imposes a condition, then its actual duration depends on how early in the round the creature acts. So, if a creature acts last and the conditions they impose expire at the end of a round, then those effects expire immediately.
If the conditions triggered during a round all take effect at the same time, then this duration problem disappears. To explain how that might work, remember that all the action in a round happens during the same six-second span—effectively simultaneously. Back in 1979, AD&D counted six seconds as simultaneous. If traded attacks came in the same segment, they hit at once, and the attackers took damage together, potentially killing both at once. AD&D segments spanned the same six seconds as fifth-edition rounds.
What if we count all the conditions triggered during a round as starting simultaneously at the beginning of the next round. Who can say when something like poison takes effect anyway? Then every effect spans the same amount of time and neatly expires or triggers a save at the end of a round. The memory demands become simpler.
More games than AD&D allowed this sort of simultaneous effect. Burning Wheel resolves every combatant’s action before applying all the results at once. Possible results include wounded, killed, unbalanced, or knocked prone. For a game like D&D, I would impose damage the moment it lands, to reduce bookkeeping. Winning initiative still gives an edge.
This feels like a big change in play, and one that not everyone can embrace. But perhaps it would play well. D&D’s designers once suspected that innovations like cyclic initiative and advantage/disadvantage might prove too big for the game, but players embraced both changes.
Dungeons & Dragons and closely related games like Pathfinder and Shadowdark use cyclic initiative where a set a turn order is set at the start of a fight and combatants cycle through the same order throughout the battle.
Initiative tents
The two best methods for tracking cyclic use cards folded into tents. Such tents enable two methods with different strengths. One technique only puts numbers on the tents, the other uses names and numbers.
To use numbers only, create a set of tents numbered from 1 up. When initiative starts, everyone compares numbers and take the card the matches their place in the order. The highest takes 1, second highest 2, and so on. The DM takes cards for the monsters’ place in the order. Everyone shows their number at their spot at the table so others can see their place. The technique always uses the same numbered tents, so it skips the need to write anything. This method doesn’t work with games like Pathfinder where a Delay action can change the initiative order.
To use names and numbers, each player puts their character name on a card. When initiative starts, the players roll and write their scores on their card. Someone collects the cards, and lines them up in initiative order where everyone can see. I drape these cards on my DM screen, but this technique also lets someone other than the game master track initiative. I delegate sorting the cards to a player.
These tracking methods make the initiative order visible to everyone. When players can see the tents and initiative order, they can see when their turn is coming and plan their actions. This speeds play. Plus, the visible initiative invites players to remind less-attentive people of their turns. It prevents GMs from accidentally skipping someone’s turn.
Pre-rolling initiative
Combat runs better when exploration or interaction flips immediately to attack rolls without the minutes of bookkeeping required to set an initiative order. To avoid postponing the action, try rolling initiative in advance, either at the end of the last fight or at the start of a session. Pre-rolling works best with names and numbers on initiative tents.
At the start of the session, while everyone unpacks their dice and chats, I typically have players pre-roll initiative for a few fights. These initiative rolls build anticipation for the session to come and fit easily in the pregame chatter. Players write their scores on initiative tents. Before the next fight starts, I delegate the task of sorting the tents. If a player wants to use Alert, the person sorting organizes the swap.
Delay adds flexibility and complexity
Unlike conferring with allies to arrange when everyone takes a turn, Ready and Delay feel like battle strategies characters might take in a split second of mayhem.
Roleplaying games with combat rules need something like the Ready action to cope with the way one creature’s turn freezes time for every other creature. Strangely, many games omit such a rule and either rely on game masters to improvise one or on players to never abuse total cover by not giving foes a sporting chance to shoot back.
Fourth edition D&D and Pathfinder also include a Delay action. In a way, this action gives players a more powerful way to tinker with the turn order than games like Draw Steel and Daggerheart, because unlike those games where the game master can intrude between two characters’ turns, Delay allows characters to coordinate actions without monsters getting turns in between. Delay also brings a price, because characters who delay fall back in initiative and keep the later place. Such tradeoffs make interesting tactical choices.
The fifth edition design team opted for a simpler game when they dropped the Delay action. The game plays fine without it, but players lose flexibility to change the turn order in a way that seems natural.
Although Delaying seems simple, it requires intricate rules. In D&D, many effects trigger at the start or end of a creature’s turn, so fourth edition needed rules summarized by this text: “You can’t Delay to avoid negative consequences that would happen on your turn or to extend beneficial effects that would end on your turn.” The fifth edition designers opted to skip all that baggage.
For an easy house rule, allow players to delay at the start of initiative before their character acts. This adds no rules complications while still creating tactical options. Delaying at the start of combat might allow the rogue to flank after the fighter moves adjacent to a foe and sets up a sneak attack.
For the players who enjoy the tactical intricacies brought by the full Delay action, groups can import the delay rules from fourth edition D&D. Here are the rules the fifth edition designers wished to avoid.
Delay
By choosing to delay, you take no action and then act normally on whatever initiative count you decide to act. When you delay, you voluntarily reduce your own initiative result for the rest of the combat. When your new, lower initiative count comes up later in the same round, you can act normally. You can specify this new initiative result or just wait until sometime later in the round and act then, thus fixing your new initiative count at that point.
You never get back the time you spend waiting to see what’s going to happen. You also can’t interrupt anyone else’s action (as you can with a readied action).
Your initiative result becomes the count on which you took the delayed action. If you come to your next action and have not yet performed an action, you don’t get to take a delayed action (though you can delay again).
If you take a delayed action in the next round, before your regular turn comes up, your initiative count rises to that new point in the order of battle, and you do not get your regular action that round.
When you Delay, any persistent damage or other negative effects that normally occur at the start or end of your turn occur immediately when you use the Delay action. Any beneficial effects that would end at any point during your turn also end. You can’t Delay to avoid negative consequences that would happen on your turn or to extend beneficial effects that would end on your turn.
Daggerheart and Draw Steel both feature advice for managing their particular versions of player-driven turn orders.
In Draw Steel, players decide which of their characters goes next. The game master chooses a monster to take a turn after each player’s turn. The rule book suggests, “To help track when creatures have already acted in the current round, each creature can have a coin, token or card they flip over on the table, or some kind of flag they set on their virtual tabletop token, once they’ve taken a turn.” (I like how Draw Steel refers to players at the table as creatures.)
This system helps, but players often forget to flip their cards. You want to see everyone’s turn status at a glance, but the scattered cards require a survey of the entire table. Also, this method does nothing to help GMs track which of their creatures have gone. As an improvement, Teos “Alphastream” Abadíacreated a GM’s screen display with cards representing each combatant. As players and monsters go, he flips down their card. “Because players will often look at the Director (GM), they see the state of battle. This worked well in play. It helped all of us have a better grasp of who was left and decide who should go.”
Tom Christy at d20Play runs games using a virtual tabletop where players can enter their initiative numbers. The VTT’s initiative tracker works with numbers, so numbers substitute for cards. He has players planning their turns enter an initiative of 0 to show that they’re unready to act. When they become ready to go, they enter 1. When they want to go immediately, they enter 2. He turns off the VTT’s automatic initiative sorting and arranges the order himself, dragging high numbers to the top to signal a turn, and then sliding creatures who act down into the next round. Tom explains his method to me in this video.
Daggerheart takes player-driven turn order further by letting players choose to allow one PC to take multiple turns in a row. For groups that prefer “structured player turns,” Daggerheart suggests players use tokens to represent the number of turns they can take, limiting everyone to three turns until everyone gets three.
Even if players choose not to limit turns this way, having a visual count of the number of turns each player takes helps show who needs spotlight time. Some game masters recommend that instead of counting down using tokens, gamers try counting up by taking tokens. Put a supply of turn tokens in the middle of the table. When players take a turn, they take a token from this pool and line it up at their place at the table so the other players can see how much time everyone has spent in the spotlight.
With player-driven turns, the biggest delays come from the moments when no one sees a reason to jump ahead of the other players. For any game with a player-driven turn order, choose a default order based on seating around the table. If no one sees an opportunity to go, just go around the table to the next person due a turn. A default turn sequence limits discussion and keeps things simple for new players and players who just want to take orderly turns. This avoids the situation where everyone tries to politely defer to the other players.
What rates as the most exciting phrase spoken in a Dungeons & Dragons game? “Roll for initiative.” What rates as the most unwelcome task? After those three words, the minutes of bookkeeping required to set the initiative order. Instead of riding the excitement of an attack, the chore drains the energy from the game.
Rather than seeking ways to minimize this delay, the 2024 D&D design team extended it. Just about every 2024 game includes a character with the Alert feat, which postpones the start of each fight with another minute of talk about who wants to swap initiative. In a wild west shootout, Black Bart reaches for his revolver, his gang raises weapons, and then the heroes take a time out to discuss who should have the quickest draw today.
To avoid stalling games just as a fight begins, some DMs have players roll initiative for the next encounter at the end of each encounter, but the Alert feat hampers this trick.
To be fair, some gamers do enjoy wringing every advantage from initiative order, but most players just want the action to start. Often, the decision of who goes next hardly matters. That can prolong the discussion as everyone politely offers the initiative to anyone else.
The trouble with talking about who goes next
The Alert feat highlights two problems with encouraging discussion about who goes next.
The extra deliberation slows play when the game should give a sense of fast action.
Talking about who goes next distracts from the game world to spotlight turn order—one of the most awkward abstractions in any RPG.
Aside from the effects of injury and the notion that everyone easily rests while spending eight hours sleeping on cold stone in a murder hole, turns rate as most unrealistic thing in D&D. The weirdness goes way deeper than how the game stops time for a discussion of who has the fastest draw today.
Turns knot time in ridiculous ways
In six seconds of actual fighting, everyone acts at the same time. But in a D&D round, turns serve as a simple but unrealistic way to make sense of six seconds. The compromise knots time in ridiculous ways. The last creature to take a turn in a round ostensibly acts in the same six seconds as the first, but typically many creatures have moved. With fifth edition’s six-second rounds, one character can end their six-second turn next to a character about to start their turn and therefore six seconds in the past. If they pass a relay baton, the baton jumps six seconds back in time. If enough characters share the same six seconds running with the baton, the object outraces a jet. Want to get the most from a Wand of Magic Missiles? Just pass it between party members and let everyone fire during the same round. Turn a Horn of Blasting into a six-second barrage!
Games like early versions of D&D, Shadowdark, and Shadow of the Weird Wizard all lack a Ready action. This simplicity exposes another awkward problem with turns: Combatants normally stay frozen in time until their turn arrives. D&D’s Blink spell only brings an advantage because foes are usually not ready to interrupt the blinker’s turn. In a chase, the distance between creatures yo-yos by 60-some feet as everyone trades turns.
Players most often exploit this unreal situation by only emerging from total cover during their turns. Imagine the party must cross a field scattered with boulders to reach a wall protected by 100 archers. In a game without a Ready action, the party can move out in plain sight, and as long as everyone ends their turn in total cover behind a boulder, then they can cross without the archers ever getting a shot. Sure, game masters can improvise a way to bring common sense, but the rules as written still fail.
Adding complexity to simulate simultaneous turns
When fourth edition D&D introduced the Ready action, D&D gained a formal rule that closed this loophole. Ready actions made turns knottier and the game more complicated, but they proved essential.
To add some of sense of turns being simultaneous, Advanced Dungeons & Dragons had people declare their actions at the start of a turn, move first, and then attack. Based on this big picture, DMs might rule that characters couldn’t reuse the same magic item during a turn, and also rule that the archers could attack when the party broke cover. But declaring actions proved cumbersome and often the changing battlefield invalidated the players’ intentions. Third edition lead designer Jonathan Tweet explains, “Eventually what you ended up doing is you had to tell the DM what you were doing every round twice.” Many tables ignored the process. Nonetheless, on the 2014 Dungeon Master’s Guide presents declaring actions as an optional initiative variant (p.270).
Who goes next
For the awkward necessity of turns to function, games need rules about who goes first and who goes next. D&D started simple. The group with the highest roll on a d6 went first. But soon these rules became complicated. First to account for things like weapon size in the name of realism, and in recent years, to emphasize tactical options or storytelling.
Modern games typically decide who goes next using one of two broad approaches:
Player-driven turn order. Games like Draw Steel, Daggerheart, and even original D&D let players decide who among the party goes next, so players can decide on the strongest order of actions. This encourages teamwork.
Cyclic turn order. Games like fifth edition D&D and Shadowdark set a turn order, and then cycle through the same order throughout the battle. This approach avoids weighing combat with ongoing decisions about who goes next, so combat moves quicker.
Side initiative
The idea of allowing players to decide the order PCs take action dates to original D&D.
Original D&D used side initiative where each group of allies took their turns together. While the player’s side has initiative, they decided how to order their character’s turns. Side initiative features the simplicity of nothing to track except who has already acted. And since players set the order for their side, they could orchestrate action combinations.
Games like Shadow of the Weird Wizard and the Cosmere Roleplaying Game use a variant of side initiative where the monsters always go first, but where players can spend one of their actions to go before the monsters. This skips an initiative roll and gives players control over when they act. The method starts each round with an engaging decision over whether to go first or to hold back and do more.
The best feature of side initiative is subtle. Side initiative (with help from early D&D’s lack of opportunity attacks) made running away from a fight much easier. When a side started their turn, they gained a chance to plan and execute an orderly retreat without any enemies interfering until everyone acted. Unlike modern D&D, where DMs typically serve fights contrived to ensure players win, early D&D’s random monsters often landed groups in deadly fights where running was the winning strategy.
To recapture some of the original game’s speed and simplicity, the fifth edition design team considered side initiative, but especially at low levels, the side that acted first gained a deadly advantage. Low level characters lack enough hit points to survive an entire round of enemy attacks. At higher levels, side initiative can turn still battles into one-sided romps when powerful spells shut down foes and attack combinations pile damage. Sure, an occasional batch of high initiative rolls can bring the same swings, but not consistently. (Side initiative appears as a variant on page 270 of the 2014 DMG.)
Players decide who goes next
Modern games with player-driven initiative typically give game masters rules for when the monsters can intrude on the turn order. So, in Daggerheart, the monsters take a turn after a player rolls with Fear. In Draw Steel, a monster goes after each player. Either way, these methods improve on side initiative by avoiding one-sided victories won because an entire side took their turns before their foes made a single action.
When players choose when characters take turns, they can make teamwork into a fun advantage where PCs get to flaunt their strengths. The caster can fireball before any allies rush into melee. The tank can rush to block charging monsters and give the rogue an opening to sneak attack. The healer can deliver a cure just in time to keep everyone fighting.
Player-driven turn orders can also foster the sort of dramatic moments common in cinema. Countless action movies set up a situation where the villain prepares a killing blow, and then gets shot when hero’s unseen ally suddenly appears. The situation counts as cliché, but we love it anyway. Systems where players can jump into initiative at any moment promote similar dramatic reversals.
Cinematic, player-driven turn order
Daggerheart takes player-driven turn order further by letting players choose to allow one PC to take multiple turns in a row. This enables the sort of sequencing common in movie battles where multiple combatants like the Avengers face off with multiple foes. The editing highlights one hero trading blows with an enemy, showing the upward beats that make for heroic moments and building tension whenever evil gains an edge (when the player rolls with Fear and the villain acts). Scenes like this rarely cut from character to character with each attack; they keep focus on a single hero until a dramatic moment prompts a cut away.
Of course, Daggerheart still plays as a game, so the optimal strategy in a fight may be to let your side’s best attacker take all the turns and make all the attacks while everyone else stands around and poses. I’ve seen movie fights like that too.
When I played fights in Daggerheart, my characters would sometimes chase foes to the edges of the map, finish them, and wind up too far away from the rest of the fight for me to feel good about asking someone to give up attacks just so I could move back. Unlike in a movie, nothing happens off camera.
Analyzing turn orders to create a narrative feels more like the judgments filmmakers make in an editing bay than like the split-second choices fighters make in combat. Perhaps the cinematic version of player-driven initiative in Daggerheart works best for players performing for an audience rather than for players making their own thrills in the moment at the table.
Weighing the merits of player-driver turn orders
How much does player-driven turn order multiply the flaws of the Alert feat by delaying the real fun of taking action? How much does it add teamwork and drama?
Player-driven turn orders add the most friction when no one sees a particular reason to jump ahead of one of their friends. Players don’t care who goes, so they act like the overly polite Goofy Gophers. “After you. No. I insist. After you.”
Player-driven turn orders play best when they enable the sort of choices a character might make in a battle: Delaying for a split second so an ally can open an advantage. Readying an attack for when a foe leaves cover.
Some of my favorite D&D sessions came when I competed in the fourth edition D&D Championship tournaments. The rules for changing turn orders offered two options: Delay and Ready. My teammates and I used those options to order turns in our favor and loved the tactical options. Besides the urgency of limited time, two factors helped us orchestrate actions without wasting time:
D&D’s cyclic initiative made changing the turn order an option rather than a constant necessity.
Before the tournaments, we practiced with the characters, so we knew the other PCs well enough to share similar opinions on who should act.
Recent games like Draw Steel and Daggerheart favor the flexibility of player-driven turn orders, but fourth edition’s take on cyclic initiative brings a better mix of play speed with decisions close to the ones combatants might make in a fight. The 13th Age roleplaying game by Jonathan Tweet and fourth edition designer Rob Heinsoo uses the fourth edition system. This game’s latest edition declares, “Jonathan introduced cyclic initiative in F20 gaming 25 years ago and we are never going back.” I understand why.
